Project SABRE was a five-year collaborative project undertaken by a multidisciplinary team from the UK, USA, and Canada, supported through the DTI Bioremediation LINK programme. Its objectives were to demonstrate that in situ enhanced anaerobic bioremediation can result in effective treatment of chlorinated solvent dense non-aqueous phase liquid (DNAPL) source areas and to improve related site investigation tools and process understanding. An important feature of the SABRE programme was the field application of DNAPL partitioning electron donor to the source zone to provide a source of electron donor at the DNAPL:water interface. The SABRE project is one of the most detailed demonstrations of its kind, and the first scientifically robust development of in situ bioremediation of a chlorinated solvent source zone in the UK.

1. CL:AIRE SABRE bulletins describe specific, practical aspects of research from the LINK Bioremediation Project SABRE, which
aimed to develop and demonstrate the effectiveness of in situ enhanced anaerobic bioremediation for the treatment of
chlorinated solvent DNAPL source areas. This bulletin describes an overview of the SABRE project, with key project topics
presented in more detail in previous and forthcoming CL:AIRE publications.
11.. IINNTTRROODDUUCCTTIIOONN
Project SABRE was a five-year collaborative project undertaken by a
multidisciplinary team from the UK, USA, and Canada, supported
through the DTI Bioremediation LINK programme. Its objectives were to
demonstrate that in situ enhanced anaerobic bioremediation can result
in effective treatment of chlorinated solvent dense non-aqueous phase
liquid (DNAPL) source areas and to improve related site investigation
tools and process understanding. An important feature of the SABRE
programme was the field application of DNAPL partitioning electron
donor to the source zone to provide a source of electron donor at the
DNAPL:water interface. The SABRE project is one of the most detailed
demonstrations of its kind, and the first scientifically robust development
of in situ bioremediation of a chlorinated solvent source zone in the UK.
This document presents an overview of the SABRE project. Key project
topics are described in previous and forthcoming CL:AIRE publications
(Table 1). The CL:AIRE and Environment Agency-funded Streamtube
project (CL:AIRE Research Project 14) ran in parallel with SABRE, at the
same site, and results from that study are reported in part in CL:AIRE
SABRE Bulletin 5 and in CL:AIRE Research Bulletin 11.
22.. TTHHEE CCHHLLOORRIINNAATTEEDD SSOOLLVVEENNTT DDNNAAPPLL PPRROOBBLLEEMM
Tetrachloroethene (PCE) and trichloroethene (TCE) are chlorinated
solvents that have been used extensively in industry and are common
groundwater contaminants (Environment Agency, 2003). Their limited
aqueous solubility often leads to the presence of DNAPL source zones at
contaminated sites that can be a long term source of groundwater
contamination. DNAPL releases may result in persistent contaminant
source areas due to mass transfer limitations from the organic phase,
and therefore low risk-based targets.
Under anaerobic conditions, PCE and TCE may be biodegraded via
reductive dechlorination (biologically-mediated step-wise replacement
of chlorine with hydrogen) to form ethene, which poses no unacceptable
risks at the concentrations arising. This microbial activity is commonly
used in the treatment of dissolved-phase TCE plumes. However,
reductive dechlorination has also been observed at concentrations
associated with the presence of DNAPL (typically >10% of aqueous
solubility), indicating that reductive dechlorination may be a viable in
situ treatment in TCE DNAPL source zones (Harkness et al., 1999). In
contrast to treatment of a dissolved phase, the primary goal of
bioenhanced DNAPL treatment is to achieve and maintain a high flux of
contaminant from the DNAPL and sorbed phases to the aqueous phase
and couple this to dechlorination near the DNAPL:water interface and
within the plume (Aulenta et al., 2006). This is achieved by increasing
the aqueous phase TCE concentration gradient close to the interface
through enhanced biodegradation of the dissolved compounds. The flux
is further enhanced because the aqueous phase solubilities of the
reductive dechlorination intermediates (cis-1,2-dichloroethene (cDCE)
and vinyl chloride (VC)) are higher than that of the parent PCE and TCE
(ITRC, 2005). Use of electron donors that partition into the DNAPL phase
and then slowly dissolve back into the water phase have been shown in
laboratory studies to provide optimum conditions for dechlorination to
occur at the DNAPL:water interface and thereby enhance dissolution
(Yang and McCarty, 2002).
33.. PPRROOJJEECCTT BBAACCKKGGRROOUUNNDD
Project SABRE was a logical incremental development from the work
performed by the US-based RTDF (Remediation Technologies
Development Forum) Bioremediation of Chlorinated Solvents
Consortium (see: http://www.rtdf.org). The consortium’s demonstration
of the effectiveness of biostimulation in treating PCE and TCE plumes at
Dover Air Force Base, Delaware (Ellis et al., 2000) and at Kelly Air Force
Base, Texas (Major et al., 2002) has led to wide uptake of those
technologies inside and outside the US. The Project SABRE team
included a number of core members of that consortium, therefore the
project benefited from the circa $15M investment in earlier applied
research. Wider UK membership, including the British Geological Survey
and several UK universities and consultancies, ensured maximum
transfer of learning to the UK and clear exploitation paths.
PPrroojjeecctt SSAABBRREE ((SSoouurrccee AArreeaa BBiiooRReemmeeddiiaattiioonn)) –– aann OOvveerrvviieeww
SSAABB 11
((SSeepptteemmbbeerr 22001100))
CCLL:AIRE
IIff yyoouu hhaavvee aannyy qquueessttiioonnss aabboouutt tthhiiss SSAABBRREE BBuulllleettiinn oorr wwoouulldd lliikkee ffuurrtthheerr iinnffoorrmmaattiioonn aabboouutt ootthheerr CCLL::AAIIRREE ppuubblliiccaattiioonnss pplleeaassee ccoonnttaacctt uuss aatt CCLL::AAIIRREE
EEmmaaiill:: eennqquuiirriieess@@ccllaaiirree..ccoo..uukk WWeebb ssiittee:: wwwwww..ccllaaiirree..ccoo..uukk
sabre bulletin
Copyright © CL:AIRE (Contaminated Land: Applications in Real Environments).
BBuulllleettiinn TTyyppee TTiittllee
Research Bulletin 6 Results of a laboratory microcosm study to determine
the potential for bioremediation of chlorinated solvent
DNAPL source areas
SABRE Bulletin 1 Project SABRE (Source Area BioRemediation) – an
overview (this document)
SABRE Bulletin 2 Site investigation techniques for DNAPL source and
plume zone characterisation
SABRE Bulletin 3 Results of laboratory column studies to determine the
potential for bioremediation of chlorinated solvent
DNAPL source areas
SABRE Bulletin 4 Insights and modelling tools for designing and improving
chlorinated solvent bioremediation applications
SABRE Bulletin 5 Overview of the SABRE field tests
SABRE Bulletin 6 Source area DNAPL bioremediation: Performance
monitoring and assessment
Research Bulletin 11 Streamtube project overview: Longitudinal transect
assessment of the SABRE site DNAPL source zone
TTaabbllee 11.. SSAABBRREE--rreellaatteedd ppuubblliiccaattiioonnss ffrroomm CCLL::AAIIRREE..

2. The LINK Bioremediation programme was founded following the Joint
Research Council Review of Bioremediation Research in the UK (1999).
One of the recommendations was the development of a “fully
integrated multidisciplinary research initiative” and in April 2001 it was
launched with the vision to “help provide UK industry with the
multidisciplinary capability necessary to enable the commercial
application of bioscience for the clean up of contaminated land, air and
water”. Over 50 outline proposals and 14 full proposals were reviewed
to arrive at a diverse portfolio of 12 projects with topics including
sequenced reactive barriers, phytoremediation, laboratory methods for
assessing bioavailability and of course both ex situ and in situ
bioremediation technologies. These projects received over £5M of
funding towards total project costs in excess of £11M over the period
2001-2009 having contributed 8 years of knowledge and experience to
the benefit of UK plc. Programme sponsors included BBSRC, BERR,
Environment Agency, EPSRC, NERC and the Scottish Government. DTI’s
LINK Bioremediation funding scheme therefore offered the SABRE
project team an opportunity to undertake a project in the UK, with
match-funding from the UK government.
The focus of the SABRE project has been the field demonstration of
enhanced bioremediation of the source zone by application of DNAPL-
partitioning electron donors. A former chemical manufacturing plant in
the East Midlands hosted the project, where a DNAPL source zone,
comprising mostly TCE, is present. Use of TCE began in 1963 and ceased
during the late 1980s.
Near-surface geology at the site, beneath Made Ground, comprises
Alluvium (1 to 2.5 m thick) overlying River Terrace Gravels (3 to 5 m
thick) over Mercia Mudstone Formation (>60 m, weathered near the
surface) (Lelliott et al., 2009).A DNAPL phase comprising approximately
71% trichloroethene (TCE) (with minor pentachloroethane, 1,2
dichloroethene and tetrachloroethene) had penetrated both of the
superficial formations and approximately the uppermost metre of the
Mercia Mudstone. Concentrations in the sediment approached 80 g/kg
(about 20% of residual saturation for this formation).Total source mass
estimates ranged from 5 to 16 tonnes. DNAPL was observed to be
heterogeneously distributed within both the Alluvium and River Terrace
Gravels. In some areas, high DNAPL saturations were observed at the
base of the River Terrace Gravels, suggesting previous migration along
the interface with the Mercia Mudstone. A plume of TCE contamination
was identified in groundwater downgradient from the source zone.
Plumes of cDCE, VC, and ethene were also observed, demonstrating
spatially variable degradation of TCE migrating from the source.
To achieve the objectives of the project, the 15 industrial and research
partners were coordinated into seven multidisciplinary work packages.
These included (1) site investigation (characterise the field site before
and after treatment), (2) laboratory research (microcosm and column
tests to determine the optimal conditions for the bioremediation
process), (3) field engineering (test cell design, installation and
operation), (4) performance assessment (traditional and novel
techniques for evaluating success), (5) flow and process numerical
modelling (development of tools to aid in design, results interpretation,
and assist future technology implementation), as well as a project
management team and dissemination committee. A Scientific Advisory
Group – consisting of three leading academics in the field – was also
appointed to help guide the project and ensure the highest scientific
standards were maintained.
44.. OOVVEERRAALLLL CCOONNCCLLUUSSIIOONNSS
44..11 DDeemmoonnssttrraattiioonn ooff tthhee TTeecchhnnoollooggyy iinn tthhee LLaabboorraattoorryy
Microcosm experiments were performed in triplicate by four laboratories
(DuPont, GE, SiREM and Terra Systems) to select the most appropriate
electron donor (energy source for the bacteria) and to determine if
nutrient addition and bioaugmentation with a TCE-degrading bacterial
culture would enhance reductive dechlorination (Figure 1). Results
showed that, in groundwater from the site, reductive dechlorination of
TCE to ethene was possible at high TCE concentrations (above 500 mg/L
in the aqueous phase) associated with a DNAPL source zone (CL:AIRE
Research Bulletin 6). Addition of an electron donor, bioaugmentation
with KB-1™ culture and addition of nutrients (diammonium phosphate
and yeast extract) all increased the rate of reductive dechlorination, and
increased the likelihood that the process would reach the ethene end-
point. Emulsified soya oil, methanol and lactate were found to be the
most effective electron donors. However, emulsified soya oil is both a
partitioning donor and has substantial practical benefits when used in
the field. Its chief practical benefit is that it can be delivered to a target
treatment zone, where it becomes immobilised and serves as a long
term source of electron donor due to its low solubility. Properly applied,
it can substantially reduce operational and maintenance costs relative to
a soluble/mobile donor. Therefore, emulsified soya oil was carried
forward to the field application.
Soil column experiments were performed by three laboratories (GE,
SiREM and DuPont). Their aims were to: understand the physical-
chemical behaviour of partitioning donors in flow systems, estimate the
extent of DNAPL dissolution enhancement created by the
bioremediation process, determine the effects of sulphate from the
Mercia Mudstone on dechlorination, and develop kinetic data to
support the modelling efforts.
The column experiments demonstrated that the emulsified soya oil
effectively partitioned into the TCE DNAPL and persisted in promoting
long-term dechlorination without causing clogging (CL:AIRE SABRE
Bulletin 3). Enhanced dissolution of the DNAPL source was observed
and DNAPL depletion rate was enhanced by a factor of approximately
two. Addition of the emulsified soya oil to columns that did not contain
TCE DNPL but that were fed high concentrations (e.g. 250 mg/L) of
aqueous phase TCE demonstrated that TCE could be completely
dechlorinated to ethene at residence times of around six days. Finally,
the column experiments demonstrated that detrimental pH shifts were
possible during very high rate reductive dechlorination but that these
shifts could be mitigated with bicarbonate addition. Results from the
SAB 1 page 2
sabre bulletin
FFiigguurree 11.. SSuubbsseett ooff tthhee mmoorree tthhaann 113300 mmiiccrrooccoossmm eexxppeerriimmeennttss ccoonndduucctteedd wwiitthh ssiittee ssooiill aanndd
ggrroouunnddwwaatteerr ((CCLL::AAIIRREE RReesseeaarrcchh BBuulllleettiinn 66))..

3. column experiments also allowed the development and calibration of
numerical models used to develop new quantitative understanding of
the dechlorination process (see Section 4.5 below).
44..22 SSoouurrccee ZZoonnee CChhaarraacctteerriissaattiioonn
The innovative “Triad” approach to site investigation was used to rapidly
assess the site condition and the locations of DNAPL sources, to
facilitate accurate detailed investigation and cell design. A three-week
long dynamic work strategy for DNAPL characterisation was established
using the following methods.
• Electrical resistivity tomography (ERT) supplied qualitative
spatial data on the topography of geological surfaces and the locations
of underground structures.
• Combined membrane interface probe (MIP) and electrical
conductivity (EC) profiling supplied qualitative data on the spatial and
vertical distribution of contamination and depths of geological
interfaces.
• Core sampling supplied quantitative data on the spatial and
vertical distribution of contamination, organic carbon, residual porosity
and the locations of geological interfaces. Sudan IV tests on cores
supplied qualitative data on the spatial and vertical distribution of
DNAPL.
CL:AIRE SABRE Bulletin 2 fully describes how qualitative and
quantitative data were used in combination to assess spatial
heterogeneity and inform the siting of the SABRE cell. To summarise,
ERT, MIP and EC data were used to guide and inform the ongoing site
investigation process as they are rapidly-deployable techniques which do
not require laboratory analyses yet still provide decision-quality data.
This allowed a potential cell location to be identified rapidly, so that
coring could be guided to the optimum locations for gathering relevant
information.
44..33 TTeesstt CCeellll CCoonnssttrruuccttiioonn,, MMoonniittoorriinngg,, aanndd TTeessttiinngg
For the field pilot test, a rectangular cell was constructed 30 m long by
3 m wide, with its long axis approximately aligned with the prevailing
hydraulic gradient (CL:AIRE SABRE Bulletin 5). The cell intersected a
DNAPL source area. Plastic sheet piles were used to construct the cell
and were keyed into the top of the Mercia Mudstone to effect
containment (this was the first time that plastic sheet piles were used in
the UK). The cell was open at the upgradient end and closed at the
downgradient end.At the influent end of the cell, a system of abstraction
and injection wells were used to homogenise, and control the rate of,
influent groundwater.At the effluent end, abstraction wells removed the
groundwater: in this way the residence time within the cell (nominally
45 days) could be controlled. Containment of the cell permitted
hydraulic control to be established over the source area so that accurate
contaminant mass balances could be computed for dissolved
contaminants.
A dense monitoring network was installed in the cell, including two
multi-level sampling fences oriented perpendicular to flow (Figure 2),
results from which were used in conjunction with point hydraulic
conductivity measurements to compute contaminant mass discharges,
decay rates and mass flux distributions (CL:AIRE SABRE Bulletin 6).
Standard fully screened monitoring wells were installed so that
“traditional” concentration and mass flux monitoring results could be
compared.Two grids of electrodes were incorporated into the multi-level
devices to allow comparison of chemical and geophysical (ERT)
monitoring techniques (Wilkinson et al., 2008). Boreholes constructed
for injection of electron donor and bioaugmentation at the source zone
were also monitored. In addition, the Streamtube monitoring network
provided further information on 2D distribution of contaminants parallel
and perpendicular to flow.
Hydraulic characterisation of the cell was performed with a combination
of tracer testing, falling head tests on each of the ports in the flux fences,
and by qualitative assessment of contaminant distribution. Hydraulic
property contrasts between the Alluvium and River Terrace Gravels were
identified early on in the programme. Some areas of the River Terrace
Gravels contained lower dissolved chlorinated solvent species
concentrations (presumably a result of enhanced flushing due to higher
permeability), and the distribution of injected electron donor was found
to vary across the test cell, suggesting that delivery was influenced by
permeability.
Even within the River Terrace Gravels, hydraulic properties showed a
highly heterogeneous 3D distribution: in the test cell hydraulic regime,
the residence time for individual flow paths varied from less than 3 days
to greater than 100 days (average 45 days). Preferential flow paths that
had been identified by the combined hydraulic characterisation
programme showed enhanced turnover of contaminants, but also hosted
more reductive dechlorination activity.
44..44 TTeesstt CCeellll OOppeerraattiioonn aanndd PPeerrffoorrmmaannccee
Cell operation commenced in January 2007 with a baseline monitoring
period that lasted approximately 100 days. The purpose of the baseline
period was to allow contaminant concentrations in the cell to come to
equilibrium under pumping conditions. Injection of SRS™ (a commercial
soya oil emulsion provided by Terra Systems Inc.) was completed in April
2007 and bio-augmentation with KB-1™ culture (provided by SiREM)
was completed in May 2007. Pumps in the cell were turned off in
January 2009 for post-remediation characterisation of the source zone.
Indications that enhanced in situ bioremediation was occurring were
identified immediately post-injection. Complete conversion of TCE to
cDCE was observed immediately down-gradient of the primary DNAPL
zone in the cell. Down-gradient concentrations of total ethenes
(TCE+cDCE+VC+ethene) began to increase after about 2.5 cell volumes
had been flushed through the cell: at this time the relative concentration
of VC increased (Figure 3). Significant increases in ethene concentration
were first reliably detected in the cell after about 3.5 cell volumes had
been flushed. SRSTM (as indicated by total organic carbon (TOC)
SAB 1 page 3
FFiigguurree 22.. SScchheemmaattiicc ooff tthhee SSAABBRREE eexxppeerriimmeennttaall tteesstt cceellll,, sshhoowwiinngg llooccaattiioonn ooff ccoonnvveennttiioonnaall mmoonniittoorriinngg
wweellllss aanndd mmuullttiilleevveell ssaammppllee ttrraannsseeccttss.. FFiigguurree ccoouurrtteessyy ooff tthhee SSttrreeaammttuubbee pprroojjeecctt..
sabre bulletin

4. measurements) appears to have become depleted near the DNAPL
source after about 5 pore volumes: this was accompanied by a decrease
in dechlorination. For future implementations careful monitoring of TOC
concentrations would indicate appropriate times for re-dosing.
Pre-and post-treatment soil data were evaluated in detail to produce
estimates of DNAPL mass removal and quantify the uncertainty in those
estimates. Upper confidence limit DNAPL mass estimates are consistent
with the removal of 875 kg ethenes in the test cell effluent and suggest
that, pre-treatment, the test cell contained in excess of 1 tonne of TCE,
of which 60-75% was removed and/or completely dechlorinated to
ethene during the field test (Figure 4).
To summarise the gains of implementing the technology over and above
natural attenuation (as monitored during the baseline period), these
were as follows (CL:AIRE SABRE Bulletin 5):
• Chlorinated ethene flux enhancement of approximately 1.6:
averaged over the life of the field test, this represented a potential
decrease in DNAPL source lifetime of 40%.
• Elimination of TCE in the treated cell effluent, and the full
conversion of 15 to 25% of the TCE to ethene or beyond in the cell.
Outside the cell, further pilot tests were undertaken to investigate
alternative implementations of the technology that might be used in the
full-scale site remediation (CL:AIRE SABRE bulletin 5). SRSTM was
compared against cheese whey, a relatively inexpensive alternative
electron donor. In terms of performance as electron donors, they were
comparable, with both causing complete dechlorination to ethene
(although SRSTM proved marginally better). However, SRSTM was found
to be economically preferable to cheese whey because the latter was
considerably more expensive to deliver to the formation and caused
formation damage by clogging (for more details see CL:AIRE SABRE
Bulletin 5).
The project developed and used five numerical models ranging from
practical tools for engineering design to research codes (Figure 5).These
numerical models have proved valuable in interpreting and
understanding laboratory and field data by providing insight into
processes (and their interactions) that are difficult to observe. In
addition, they are valuable for designing bioremediation schemes and
optimising their performance. At all scales and degrees of complexity,
models are particularly useful for evaluating what-if scenarios and
examining the sensitivity of bioremediation performance to the natural
and engineered site conditions. A summary of the modelling tools
developed, including their potential to support and optimise future
remediation scheme design, is outlined below.
A groundwater flow model was developed in MODFLOW to analyse the
hydraulic conditions in the SABRE field test cell and to evaluate different
electron donor injection strategies with the objective of maximising its
distribution while minimising the on-site application cost.
COMPSIM is a comprehensive numerical groundwater fate and transport
model capable of simulating multi-phase, multi-dimensional and multi-
component systems (Sleep and Sykes, 1993). This model provides a
practical and computationally efficient modelling platform for simulating
dechlorination at the field scale. For situations where complex hydraulic
or geochemical field conditions prevail, this model may be a useful
design tool for engineers that, once calibrated for those field conditions,
may be applied by the developers for interested parties.
BUCHLORAC was created as a practical and easy-to-use software tool
that can provide detailed buffer dosage estimates for field
dechlorination projects. With many remediation sites prone to
groundwater acidification as dechlorination proceeds, implementation of
pH control strategies may be crucial to the design of successful
treatment schemes. The program is available as free supplementary
material with Robinson and Barry (2009) or from
http://infoscience.epfl.ch/record/135054.
BUCHLORAC_FLOW, an extension of BUCHLORAC that includes flow
and reactive transport in multiple dimensions, was able to demonstrate
how the acidity build up is influenced by the field-scale flow system,
including in situ heterogeneity (Brovelli et al., 2010). This model is
primarily a research tool, but could be readily applied to field sites by the
developers on behalf of interested parties.
BIOPROCESS is a comprehensive model designed to simulate the
complex suite of physical, biological, and geochemical processes
interacting in bioremediation (Kouznetsova et al., 2010).This model was
able to provide significant insight into the feedbacks between processes
SAB 1 page 4
FFiigguurree 44.. PPrree-- aanndd ppoosstt--ttrreeaattmmeenntt DDNNAAPPLL ddiissttrriibbuuttiioonn iinn tthhee tteesstt cceellll.. RReeddss aanndd yyeelllloowwss rreepprreesseenntt
hhiigghheerr rreessiidduuaall ssaattuurraattiioonn ((CCLL::AAIIRREE SSAABBRREE bbuulllleettiinn 55))..
FFiigguurree 33.. CChhlloorriinnaatteedd eetthheenneess iinn tteesstt cceellll eefffflluueenntt..
sabre bulletin
FFiigguurree 55.. FFiivvee nnuummeerriiccaall mmooddeellss ccrreeaatteedd//mmooddiiffiieedd wwiitthhiinn SSAABBRREE ffoorr aaddddrreessssiinngg aassppeeccttss ooff iinn ssiittuu
eennhhaanncceedd bbiioorreemmeeddiiaattiioonn ((CCLL::AAIIRREE SSAABBRREE bbuulllleettiinn 44))..

5. controlling effective bioremediation design, including sensitivity to
potential inhibition factors. It also assisted in analysis of the column
experimental data and the understanding gained from this model will
help direct future remediation scheme design. However, due to the
model's complexity, high computational demand and input
requirements, its primary use is further research.
The SABRE modelling program is described in detail in CL:AIRE SABRE
Bulletin 4.
55.. AAPPPPLLIICCAATTIIOONN OOFF TTHHEE TTEECCHHNNOOLLOOGGYY
Two sets of issues should be considered when considering the
application of this technology in the field: environmental (applicability of
the technology given differing geochemical and hydrogeological
conditions) and management (likely acceptability to regulators and site
management, including factors such as cost, project time constraints,
reliability and sustainability).
55..11 EEnnvviirroonnmmeennttaall IIssssuueess:: GGeeoocchheemmiissttrryy
Numerous processes competed for the injected organic substrate.
Sulphate reduction, iron reduction and methanogenesis all consumed a
proportion that could not contribute to dechlorination. In the test cell,
methanogenesis was largely inhibited due to the high TCE
concentrations, making the dechlorination process much more efficient.
Sulphate reduction was more important because of in-flow of sulphate-
rich groundwater from the Mercia Mudstone beneath the cell. Slow
fermentation of the SRSTM organic substrate provided a long term source
of carbon for the most efficient dechlorination.
As in the column studies, pH excursions were observed in the field after
about one pore volume had been pumped through the test cell: where
pH values began dropping towards pH 6 as the reductive dechlorination
reaction accelerated. Due to concerns that microbial activity would be
compromised, influent groundwater was dosed with potassium and
sodium bicarbonate and the pH quickly stabilised to more neutral values.
Some aquifer plugging was noted but could not be explicitly tied to the
bicarbonate addition.
Competitive and toxic inhibition of the dechlorination process is
important. For example, at TCE concentrations approaching 500 mg/l the
dechlorination reaction rate is reduced by 85% due to the high
concentration. However, consistent (but slow) dechlorination is predicted
at concentrations up to 750 mg/l TCE (CL:AIRE Research Bulletin 6).
55..22 EEnnvviirroonnmmeennttaall IIssssuueess:: HHyyddrrooggeeoollooggyy
The SABRE project has demonstrated that this technology is likely to be
suitable for use in many geological and hydrogeological conditions but
with some important qualifications. Groundwater flow rate (i.e.
residence time in the active treatment zone) is an important factor in
controlling the process. Higher flow rates decrease the overall time
available for conversion of TCE to ethene and can result in incomplete
dechlorination.
The technology is most suitable for use in sandy aquifer settings. It may
be applied at relatively high groundwater velocities through careful
selection of electron donor and donor delivery method. For example,
where emulsified vegetable oil (and similar products) adsorbs to the
aquifer matrix (porous media and fracture walls) and partitions into the
DNAPL. This allows the technology to function effectively, even in highly
permeable environments where soluble electron donors, such as
molasses and lactate, would be rapidly flushed out of the treatment
zone. However, the dominance of preferential flow paths in fractured
aquifers suggests that the technology may be less suitable under such
conditions. The technology is unlikely to be effective in low-permeability
strata, where amendment delivery would be problematic.
Although the average residence time in the SABRE test cell was 45 days,
the residence time for individual flow paths varied over two orders of
magnitude. Complete dechlorination occurred within the longer flow
paths, but not along the shorter ones, resulting in test cell effluent that
contained a mixture of TCE and its daughter products including ethene.
In full-scale application, similar behaviour is expected although there
could, of course, be a larger total span in residence time; and the longest
residence times would typically be associated with the smallest flux.
Adequate residence time must therefore be provided for complete
dechlorination.
55..33 MMaannaaggeemmeenntt IIssssuueess
Under many circumstances in situ bioremediation may be the lowest cost
and most sustainable remedial option, making it the preferred
technology. Its extended treatment times may not be a disadvantage at
operational sites, while reduced infrastructure may also be highly
favourable. The technology can also be safely applied near or under
existing buildings, although man-made preferential flow paths may lead
to complicated flow patterns.
Although this technology has the potential to decrease remediation
timeframes relative to natural attenuation or pump and treat, mass
transfer limitations out of DNAPL source zones can still result in
extended timeframes, depending on remedial targets. The relatively long
treatment timescales and potential cost uncertainties associated with in
situ bioremediation could result in it being unsuitable (when used alone)
for redevelopment projects or in other circumstances where completion
of remediation within a short time-frame may be required.
Table 2 summarises part of a comparative technology evaluation carried
out for full-scale remediation at the SABRE research site. This illustrates
the relatively lower cost and high sustainability of in situ bioremediation
for treatment of DNAPL source zones under site conditions. While this is
a site-specific assessment, the results may be generalised.
55..44 AApppplliiccaattiioonn ooff tthhee TTeecchhnnoollooggyy iinn tthhee UUKK
For PCE and TCE source zone remediation the use of in situ
bioremediation can be considered as part of the CLR11 options appraisal
process (Environment Agency, 2004). Its use is likely to be preferable on
sites where the efficacy and benefits through more sustainable
remediation can be demonstrated. The aim of the Environment Agency
in sponsoring the SABRE research has always been to promote improved
remedial methods.
In situ enhanced bioremediation for DNAPL source areas using
emulsified vegetable oil has already been implemented successfully at
full-scale at a site in the UK, winning the Brownfield Briefing Award
2009 for best use of a single remediation treatment technique.
A second phase of emulsified vegetable oil injection in the test cell was
carried out in March 2010. Full-scale treatment at the SABRE research
site is scheduled to start in late 2010.
SAB 1 page 5
sabre bulletin

6. AAcckknnoowwlleeddggmmeennttss
The SABRE project team comprises (in no particular order): Acetate
Products, Akzo Nobel, Archon Environmental, British Geological Survey,
Chevron, DuPont, ESI, General Electric, Geosyntec Consultants, Golder
Associates, Honeywell, Scientifics, Shell Global Solutions, Terra Systems,
University of Edinburgh, University of Sheffield and CL:AIRE.
The SABRE project team would like to acknowledge the in-kind and
direct financial contributions made by each participating organisation.
SABRE was funded through the DTI Bioremediation LINK programme
(http://www.clarrc.ed.ac.uk/link/) with specific financial contributions
from BBSRC, DTI, the Environment Agency, EPSRC and NERC, which are
gratefully acknowledged.
This Bulletin was prepared by members of the Project SABRE team,
specifically: Steve Buss (ESI), Colin Cunningham (LINK), David Ellis
(DuPont), Ian Farrar (ESI), Jason Gerhard (University of Edinburgh, now
at University of Western Ontario), Mark Harkness (GE), Alwyn Hart (EA),
Lawrence Houlden (Archon Environmental), Mike Lee (Terra Systems),
Gary Wealthall (BGS), Ryan Wilson (University of Sheffield) and Peter
Zeeb (Geosyntec).
RReeffeerreenncceess
Aulenta, F., Majone, M. and Tandoi, V., 2006. Enhanced anaerobic bioremediation of chlorinated
solvents: environmental factors influencing microbial activity and their relevance under field
conditions. Journal of Chemical Technology & Biotechnology, 81, 1463-1474.
Brovelli, A., Barry, D.A., Robinson, C. and Gerhard, J.I., 2010. Field scale modeling of enhanced
reductive dechlorination, acidity production and remediation failure. Submitted to Advances in
Water Resources.
Environment Agency, 2003. Illustrated handbook of DNAPL transport and fate in the subsurface.
Environment Agency R&D Publication 133.
Environment Agency, 2004. Model Procedures for the Management of Land Contamination.
Environment Agency Contaminated Land Report 11.
Ellis, D.E., Lutz, E.J., Odom, J.M., Buchanan Jr., R.J., Lee, M.D., Bartlett, C.L., Harkness, M.R. and
Deweerd, K.A., 2000. Bioaugmentation for accelerated in situ anaerobic bioremediation.
Environmental Science and Technology, 34, 2254-2260.
Harkness, M.R., Bracco, A.A., Brennan Jr., M.J., DeWeerd, K.A. and Spivack, J.L., 1999. Use of
bioaugmentation to stimulate complete reductive dechlorination of trichloroethene in Dover soil
columns. Environmental Science and Technology, 33, 1100-1109.
ITRC (Interstate Technology & Regulatory Council), 2005. In situ bioremediation of chlorinated
ethene DNAPL source zones. BioDNAPL-1. www.itrcweb.org
Kouznetsova I., Mao, X., Robinson, C., Barry, D.A., Gerhard J.I. and McCarty P.L., 2010. Biological
reduction of chlorinated solvents: Batch-scale geochemical modeling. Advances in Water
Resources, in press, doi: 10.1016/j.advwatres.2010.04.017.
Lelliott, M., Cave, M. and Wealthall, G., 2009. A structured approach to the measurement of
uncertainty in 3D geological models. Quarterly Journal of Engineering Geology and Hydrogeology,
42, 95-105.
Major, D.W., McMaster, M.L., Cox, E.E., Edwards, E.A., Dworatzek, S.M., Hendrickson, E.R., Starr,
M.G., Payne, J.A. and Buonamici, L.W., 2002. Field demonstration of successful bioaugmentation
to achieve dechlorination of tetrachloroethene to ethene. Environmental Science and Technology,
36, 5106-5116.
Robinson, C. and Barry, D.A., 2009. Design tool for estimation of buffer requirement for enhanced
reductive dechlorination of chlorinated solvents in groundwater. Environmental Modelling and
Software, 24, 1332-1338.
Sleep B.E. and Sykes, J.F., 1993. Compositional simulation of groundwater contamination by
organic compounds. 1. Model development and verification. Water Resources Research, 29, 1697-
1708.
Wilkinson, P., Chambers, J., Lelliott, M., Wealthall, G. and Ogilvy, R., 2008. Extreme sensitivity of
crosshole electrical resistivity tomography measurements to geometric errors. Geophysical Journal
International, 173, 49-62.
Yang, Y. and McCarty, P.L., 2002. Comparison between donor substrates for biologically enhanced
tetrachloroethene DNAPL dissolution. Environmental Science and Technology, 36, 3400-3404.
SAB 1 page 6
This document is printed on Era Silk recycled paper (FSC certified TT-COC-2109) using vegetable-based inks
TTeecchhnnoollooggyy SSAABBRREE aapppprrooaacchh
iinn ssiittuu bbiioorreemmeeddiiaattiioonn
EExxccaavvaattiioonn aanndd ooffff--ssiittee ddiissppoossaall HHiigghh tteemmppeerraattuurree iinn ssiittuu tthheerrmmaall
ttrreeaattmmeenntt
IInn ssiittuu cchheemmiiccaall ooxxiiddaattiioonn
Description of
technology
In situ bioremediation using
reductive dechlorination
enhanced by amendment with
emulsified vegetable oil (EVO),
soluble electron donor, nutrients,
together with bioaugmentation
if appropriate
Excavation of source to 7 m depth
inside a tent with full air
containment and off gas treatment.
Excavation requires dewatering and
treatment of abstracted
groundwater. Off-site disposal to
landfill including UK Landfill Tax
In situ high temperature thermal
treatment using steam injection
In situ chemical oxidation using
potassium permanganate or
Fenton's Reagent
Total costa £1.5 million £2.5 million £2.85 million £3.0 million
Unit cost of soil
treatment, £ per m3
£300 per m3 £500 per m3 £570 per m3 £600 per m3
Confidence in cost
estimate
Moderate: estimate based on
site-specific treatability data,
pilot trials and unit rates
High: market-tested cost based on
competitive tender against defined
contractual terms
Moderate to High: market-tested
quotation from contractor but
contract terms not fully defined
Moderate to High: market-
tested quotation from
contractor but contract terms
not fully defined
Timescale 3 to 5 years 6 months 15 months 1.5 to 2 years
Confidence in
timescale
Low High Moderate Moderate
Major advantages Most sustainable solution
Potentially lowest cost
Can be used safely close to and
beneath site infrastructureb
Rapid
Moderate cost
Cost certainty
Short treatment time Short treatment time
May be safe to use close to and
beneath buildings and other
site infrastructure
Major disadvantages Cost uncertainty
Long timescale
Poor sustainabilityc
Risk of fugitive emissionse
Not suitable for remediation
beneath site infrastructureb unless
these can be removed prior to
treatment
Poor sustainability score due to
large energy use
Risk of fugitive emissions of VOCs
Risk of ground heave causing
damage to site infrastructureb
Likely higher cost.
Moderate sustainabilityc due to
chemical use
Aquifer foulingd
Use of hazardous chemicalsf
TTaabbllee 22.. CCoommppaarriissoonn ooff tteecchhnnoollooggyy ooppttiioonnss ffoorr ffuullll--ssccaallee rreemmeeddiiaattiioonn ooff DDNNAAPPLL ssoouurrccee aarreeaa aatt tthhee SSAABBRREE rreesseeaarrcchh ssiittee..
sabre bulletin
a Independent supervision and 12 month validation monitoring included in all options.
b Infrastructure includes buildings, site roads, car parks and underground services.
c Sustainability: Excavation and off-site disposal scored low due to (i) transport impacts from large-
scale transportation of soil and (ii) contaminants are transferred to contained landfill not treated.
d Associated with potassium permanganate due to precipitation of manganese dioxide.
e From excavations, stockpiles and vehicle movements.
f Fenton's Reagent requires use of hazardous chemicals (acid and hydrogen peroxide).